chore: merge main into add-screen-reader branch #726
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
* chore: rearrange html to be a little nicer * fix: dont show shortcuts with no keys * fix: update shortcuts shown in dialog * fix: i18n for keyboard help header * fix: fix return type
Fixes #686 This updates the GH pages workflow to: - Automatically deploy with changes to both `main` and `add-screen-reader-support-experimental` (the screen reader experimental branch--see RaspberryPiFoundation/blockly#9283). - Upload the screen reader version of the playground to a new `/screenreader` subdirectory in the plugin repo's GitHub pages deployment. This will approximately double the amount of time needed for the GH pages deployment since it requires building two distinct copies of core Blockly and the plugin (tip-of-tree versions and the screen reader experimentation versions). Conveniently, this will act as a fast way to see screen reader changes since they will go live a few minutes after a change to the plugin. Unfortunately, that does not reciprocate for changes to core Blockly (where most of the screen reader changes will actually occur), so oftentimes a manual run of the workflow will be needed (which should suffice in those cases). Note that this needs to be approached in this way (building and uploading both projects for changes to either branch) since the action being used for deployment doesn't support partial deployment. See: actions/deploy-pages#349.
…dates (#683) Bumps the npm_and_yarn group with 2 updates in the / directory: [on-headers](https://github.com/jshttp/on-headers) and [compression](https://github.com/expressjs/compression). Updates `on-headers` from 1.0.2 to 1.1.0 - [Release notes](https://github.com/jshttp/on-headers/releases) - [Changelog](https://github.com/jshttp/on-headers/blob/master/HISTORY.md) - [Commits](jshttp/on-headers@v1.0.2...v1.1.0) Updates `compression` from 1.8.0 to 1.8.1 - [Release notes](https://github.com/expressjs/compression/releases) - [Changelog](https://github.com/expressjs/compression/blob/master/HISTORY.md) - [Commits](expressjs/compression@1.8.0...v1.8.1) --- updated-dependencies: - dependency-name: on-headers dependency-version: 1.1.0 dependency-type: indirect dependency-group: npm_and_yarn - dependency-name: compression dependency-version: 1.8.1 dependency-type: indirect dependency-group: npm_and_yarn ... Signed-off-by: dependabot[bot] <[email protected]> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: RoboErikG <[email protected]>
* fix(tests): Restore window size after scroll tests * docs(tests): "Selenium" -> "WebdriverIO" We aren't using Selenium, so we shouldn't be saying we do. * chore(tests): Make timeout(0) calls only when PAUSE_TIME is nonzero With PAUSE_TIME set to zero, on my relatively fast laptop none of these tests run slow enough to fail due to the default 2s timeous, and even with PAUSE_TIME set to 50ms the slowest test is 1024ms (of which 700ms is spent in brower.pause calls inside sendKeyAndWait). We do want to disable timeouts when debugging tests, both because we might set PAUSE_TIME to a larger value and because we might use the debugger to pause things entirely, but we don't need to disable timeouts when running tests just to check their results. * fix(tests): Set PAUSE_TIME to 0 It's very useful to be able to make tests run more slowly so you can watch them, but they should run as fast as possible by default. This cuts total test execution time on a 2021 MacBook Pro M1 approximately in half, from 42s to 22s. * refactor(tests): Use sendKeyAndWait where appropriate In most places we were already following browser.keys with a browser.pause(PAUSE_TIME), so using sendKeysAndWait is more succinct; in other places we didn't have the pause but it is not harmful to add a pause (especially now the default PAUSE_TIME is zero) and could be helpful when watching tests run with a non-zero PAUSE_TIME. * fix(tests): Make tabNavigateToWorkspace idempotent Previously this function would just send a bunch of tabs, which depended on focus state being as-on-document-load. Some tests (notably the ones in basic_test.ts) that have only a suiteSetup and not a (per-test) setup method were only were only passing because of the combination of: * Due to issue #632, pressing tab when the workspace is focused (and there are no further focusable elements on the page) incorrectly causes focus to move to the first focusable element on the page instead of (as would normally be the case) to the browser controls, and * The fact that the index.html had exactly one additional focusable div on the page, preceding the injection div. This meant that calling tabNavigateToWorkspace when the workspace was already focused would, only by sheer coincidence, result in the focus remaining on the workspace. By explicitly focusing a known element, tabNavigateToWorkspace should work correctly regardless of how many focusable elements are on the page and which one (if any) was focused before the call. * chore(tests): Remove unneeded tabNavigateToWorkspace calls Any time a tabNavigateToWorkspace call is followed by a call to focusOnBlock the former can be removed with no discernable effect except to make tests run slightly faster and with less flashing of the flyout. * fix(tests): Increase timeout for certain slow tests * docs(tests): Fix typo * chore(tests): Add missing timeout(0) calls (when PAUSE_TIME is nonzero) These two files were inadvertently omitted from commit 14d619c. * fix(tests): Lint * fix(tests): Add missing import
* chore: Added context menu item tests. * Maybe clicking a different location fixes the github test failure. * Assigned the expected actions to constants at top of context menu test. * Fix shadow block action test in github action. * Fix shadow block action test in github action attempt 2.
The focusOnBlock function moves focus to the workspace regardless of where it was previously, so there's no need to call tabNavigateToWorkspace first.
* test(MoveActions): Test starting move using keyboard shortcuts Fixes #695. * refactor(tests): Make sendKeyAndWait faster if PAUSE_TIME === 0 Making a single webdriverio call is much faster when sending repeating keypresses; do that if we don't need to pause.
* refactor(tests): Rename move test blocks to move start test blocks Also remove an unnecessary round trip through JSON.stringify + JSON.parse. * test(Mover): Introduce test blocks and suite for statement move tests Introduce moveStatementTestBlocks and suite 'Statement move tests', for testing movement of statement blocks. Move the existing test for constrained movement following unconstrained movement to the new suite. * test(Mover): Test constrained move of simple stack block left/right * refactor(tests): Refactor simple stack block move left/right Introduce a new helper function, testMove, that encapsulates the work needed to test moving a block through a constrained move, checking candidate connections and final position. Use this helper to reimplement the move left/right test as two separate tests. * test(Mover): Test constrained move of simple stack block up/down * docs(tests): Improve documentation of getFocusedNeighbourInfo * test(Mover): Also check index of candidateConnection.local Modify getConnectionCandidate to also return the index of candidateConnection.local in the list of connections returned from block.getConnections_(true), so we can check wich connection on the moving block is being used to connect. For the simple mover case this is not very interesting (it's normally index 0, for the previous connection, occasionally index 1 if the moving block will be the new top block in a stack) but this will be more important for tests involving moving more complexly-shaped blocks. * chore(tests): Add block with statment inputs to statement test blocks Add a new, E-shaped block with ID complex_mover to moveStatementTestBlocks, and update expected test results for statement move tests. * test(Mover): Test constrained move of stack block with statement inputs Note that this test verifies current behaviour which, due to bug #702, does not conform to the desired behaviour. There are TODOs to update tests to desired behaviour when bug is fixed. * docs(tests): Correct copied comment
* fix(tests): Suppress tsc unintentional comparison error
When PAUSE_TIME is set to a value other than 0 the comparison
in sendKeyAndWait generates a TS unintentionl comparison error.
* refactor(tests): Simplify moveTest
Simplify moveTest (and the tests that use it) by assuming that
the move always finishes in the same location it started in.
We will leave testing various move end scenarios (e.g. healing)
for a separate suite of move-finishing tests, akin to existing
move start tests.
* fix(tests): eslint-disable to ignore @ts-igonre
* test(Mover): Introduce test blocks and suite for value move tests
Introduce moveValueTestBlocks and a suite 'Value expression move
tests' for each of two renderers (geras, zelos) to test constrained
movement of value blocks.
* test(Mover): Add value expression right/left move tests
Moving value blocks left/right behaves as expected when using
geras, but in zelos there is a bug where external inputs
are not visited in the expected order (not that external inputs
_look_ external when using geras, but still it is a possible
configuration and should behave as expected).
Note that two of these tests fail due to a bug; this will be dealt
with in the next commit.
* refactor(tests): Use text_join, nested suites
Refactor the simple value expression move tests as follows:
- Use text_join rather than text_count, as the former is a more
plausible example of an value block with multiple external
inputs.
- Use common outer suite with nested suites for each of the
different renderers, moving common constants to outer suite.
* feat(tests): Add blocks for complex value block move tests
* refactor(tests): Further organise with nested suites
* test(Mover): Add value row (no free inputs) right/left tests
* test(Mover): Add unary expression right/left tests
* chore(tests): Skip failing zelos value block move tests due to #707.
* chore(tests): Skip failing unary move right tests due to #709
* fix(tests): Remove debugging console.log
* fix(tests): Ignore unhelpful lint error
* chore: Update Blockly dependency to 12.3.0. * release: bump version to 3.0.0 * chore: Update Blockly peer dependency to 12.3.0. * release: bump version to 3.0.1
… move tests (#711) * feat(tests): Introduce appendBlock test helper Creaet a new test helper appendBlock, that will create a new block from JSON and connect it as requested. * refactor(tests): Use appendBlock in statement move tests This means the test blocks don't need to include the block(s) to be moved. * refactor(tests): Use appendBlock in value move tests This means the test blocks don't need to include the block(s) to be moved, which considerably simplifies adding several more test cases. * chore(tests): Improve value move tests checking vs. unconnected block Use an unconnected (top-level) block with two empty value inputs, to allow us to better compare order of visiting inputs vs. outputs of top-level blocks. * test(Mover): Add complex expression right/left move tests
* chore: Make the default test page include advanced playground features. * chore: Remove unused import. * fix: Fix registration exceptions on reinjection.
* chore: fix a bug in flaky move test * chore: add assert to make sure context menu is selected * chore: try using keyboard to start move * chore: lint * Revert "chore: lint" This reverts commit c906f12. * Revert "chore: try using keyboard to start move" This reverts commit 6a474f7. * fix: add class to move indicator bubble * chore: minor refactor
maribethb
temporarily deployed
to
github-pages
GitHub Actions
Inactive
RoboErikG
approved these changes
Sep 12, 2025
Contributor
RoboErikG
left a comment
RoboErikG
left a comment
There was a problem hiding this comment.
It looks like the workflows were changed to run against main instead of the local branch. Should that piece be reverted?
| push: | ||
| branches: | ||
| - add-screen-reader-support-experimental | ||
| - main |
Contributor
There was a problem hiding this comment.
Should the branch names get changed?
cpcallen
temporarily deployed
to
github-pages
GitHub Actions
Inactive
Collaborator
Author
Thanks, this is fixed now that Christopher updated main again. |
gonfunko
temporarily deployed
to
github-pages
GitHub Actions
Inactive
maribethb
merged commit a86068b
into
add-screen-reader-support-experimental
14 of 16 checks passed
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Merges main in to keep feature branch up to date and pull in new features & fixes.